Viremic HIV Controllers Exhibit High Plasmacytoid Dendritic Cell–Reactive Opsonophagocytic IgG Antibody Responses against HIV-1 p24 Associated with Greater Antibody Isotype Diversification

Combination antiretroviral therapy (ART) is extremely effective in controlling HIV replication but cannot eradicate the infection. HIV genomes integrate into DNA of long-lived cells, such as central memory CD4⁺ T cells, and form a latent reservoir of infection that reactivates when ART is ceased. Furthermore, individuals with HIV infection treated with ART may experience low-level viral replication, which contributes to immune activation, inflammation, and activation of the coagulation system that are associated with an increased risk of atherosclerotic vascular disease, osteoporosis, and non-AIDS cancers (1). A large international research effort is currently focused on ways to decrease the size of latent HIV reservoirs and potentially eradicate the infection (2). It is generally accepted that the initial step should be to activate the reservoir of HIV proviral DNA from latency with latency inhibitors such as histone deacetylase inhibitors (3). However, inhibiting HIV latency alone is unlikely to decrease the size of the HIV reservoir and other measures, such as enhancement of endogenous retroviral restriction factors and/or “protective” immune responses against HIV Ags by therapeutic vaccines, are likely to be required to eliminate HIV-infected cells (4). It is therefore important to elucidate protective immune responses against HIV that have the potential to be enhanced by a therapeutic vaccine.

Data from numerous studies of individuals who can naturally control HIV infection (HIV controllers) indicate that the strongest correlate of immune control is CD8⁺ T cell responses against proteins encoded by the Gag gene of HIV that are restricted by particular protective HLA-B alleles, especially HLA-B*57 (5). Peptides of HIV Gag proteins are expressed by class I MHC molecules of T cells latently infected by HIV (6, 7) and are potential targets for vaccine-induced immune responses. However, vaccines that induce T cell responses against HIV Gag proteins have been ineffective in preventing or controlling HIV infection (8). Research efforts are therefore being focused on enhancing other protective immune responses. Studies in simian HIV–infected macaques have shown that human mAbs against HIV-1 Env Ags suppress replication of simian HIV and are capable of inducing long-term suppression of simian HIV infection in a subset of animals (9, 10). Numerous studies have also demonstrated that IgG Abs against HIV-1 Gag proteins are associated with slower progression of HIV disease (reviewed in Ref. 11) but it is unclear what role, if any, these Abs play in controlling HIV-1 replication.

Studies of acute SIV infection in macaques have shown that IFN-α suppresses SIV replication, although prolonged exposure to IFN-α has deleterious effects (12). Additionally, administration of IFN-α therapy to HIV patients receiving ART may decrease the size of the HIV DNA reservoir (13, 14). Natural control of HIV-1 replication is associated with higher activity of IFN-α–stimulated NK cells (15) and of plasmacytoid dendritic cells (pDCs) (16, 17), which are the major producers of IFN-α. pDCs can be activated to produce IFN-α by opsonophagocytic Ab responses against coxsackieviruses (18) and picornaviruses (19), both of which are nonenveloped RNA viruses. Therefore, although a mechanism for control of HIV replication by IgG Abs against Gag proteins remains to be established, it is possible that they exert opsonophagocytic Ab activity that activates pDCs to produce IFN-α.

Opsonophagocytic Ab responses against encapsulated bacteria are enriched for IgG2 Abs against capsular polysaccharides (20–22). This is likely to reflect structural and functional characteristics of IgG2 Abs that promote opsonization of multivalent Ags and preferential binding to FcRs that facilitate phagocytosis. Thus, IgG2 Abs can exist as multiple structural isoforms consisting of different H chain/L chain/hinge covalent complexes that result from differences in disulfide bonding between the hinge and Fab regions of the molecule (23–25) and also form covalent dimers at the hinge region (26, 27). Additionally, whereas IgG2 Abs bind to FcRs with lower affinity than other IgG subclasses, binding is predominantly to the 131H genotype of FcγRIIa (28), which is carried by 75% of individuals from all racial groups and is the major FcγR mediating phagocytosis (29). Furthermore, IgG2 and IgG1 Abs are more effective than IgG3 Abs in activating intracellular pathways of DCs following phagocytosis via FcγRIIa (30). FcγRIIa is expressed on multiple cell types that undertake Ab-dependent phagocytosis, including pDCs on which FcγRIIa is the most abundant activatory FcγR (31). Of note, the 131H genotype of FcγRIIa is associated with slower progression of HIV disease (32).

An IgG2 Ab response against HIV proteins has been associated with control of HIV replication or slower HIV disease progression (33, 34), and we have provided evidence that Abs against HIV Gag proteins that include IgG2 Abs may contribute to HIV immune control mechanisms (35). There is also evidence that IgG2 Abs may be beneficial in controlling other viruses. For example, IgG2 Abs against L1 capsids of human papillomavirus-16 have been associated with regression of human papillomavirus-16⁺ cervical intraepithelial neoplasia (36). Production of IgG2 Abs during the maturation of an Ab response occurs through class switch recombination of IgH genes, in which the “upstream” Cγ3 (IgG3) and Cγ1 (IgG1) genes of the IgH locus are switched with the Cγ2 (IgG2) and Cγ4 (IgG4) genes located farther downstream (37, 38). Thus, switching to “downstream isotypes,” to include IgG2, leads to isotype diversification of the IgG Ab response and broadens Ab function.

In this study, we have examined IgG Ab responses against HIV-1 p24 in HIV controllers and noncontrollers and demonstrated that both pDC-reactive opsonophagocytic Ab responses and greater isotype diversification are associated with control of HIV-1 infection in patients with viremic HIV infection.

Cryopreserved plasma samples were obtained from two groups of HIV patients: 1) 34 ART-naive adult HIV patients recruited in Perth, consisting of 14 HIV controllers (plasma HIV RNA levels < 2000 copies/ml for at least 1 y) and 20 HIV noncontrollers (plasma HIV RNA levels > 5000 copies/ml) (Perth cohort), and 2) 89 ART-naive adult HIV patients from the University of California San Francisco SCOPE cohort, consisting of 30 elite HIV controllers (HIV RNA levels < 75 copies/ml for at least 1 y), 29 viremic HIV controllers (HIV RNA levels 75–2000 copies/ml for at least 1 y), and 30 HIV noncontrollers (HIV RNA levels > 10,000 copies/ml). Demographic characteristics for the two study groups are summarized in Table I. Informed consent was obtained from all subjects and the study was approved by the Ethics Committees of Royal Perth Hospital and the University of California San Francisco.

Three cell lines were used for the assays of opsonophagocytic Abs. The Gen2.2 (pDC) cell line was obtained from the French Collection of National Microorganism Cultures (Institut Pasteur, Paris, France) and was described previously by Chaperot et al. (39). Gen2.2 cells were grown in a 75-cm² culture flask (Cellstar, Monroe, NC) with an 80% confluent MS-5 feeder monolayer, 20 ml RPMI 1640 medium supplemented with GlutaMAX (Life Technologies, Mount Waverely, VIC, Australia), and 10% heat-inactivated FBS (Life Technologies). Gen2.2 cells (4 × 10⁶) were passaged every third day into a new 75-cm² culture flask containing a fresh 80% confluent MS-5 feeder monolayer and 20 ml Gen2.2 growth media. The second cell line, the irradiated murine stromal cell line MS-5, was obtained from the German Collection of Microorganisms and Cell Cultures (Leibniz Institute Deustsche Sammlung von Mikroorganismen und Zellkulturen, Braunschweig, Germany). Adherent MS-5 feeder cells were grown in 20 ml α-MEM (Life Technologies) supplemented with 10% heat-inactivated FBS in a 75-cm² culture flask. MS-5 cells were detached with 0.25% trypsin/EDTA (Life Technologies) and passaged 1:3 every third day into a new 75-cm² culture flask containing 20 ml fresh MS-5 growth media. The third cell line, the THP-1 (monocyte) cell line, was obtained from laboratory stocks. THP-1 cells were grown in 30 ml RPMI 1640 supplemented with 2 mM l-glutamine (Life Technologies) and 10% heat-inactivated FBS. THP-1 (4 × 10⁶) cells were passaged into a 75-cm² culture flask containing 30 ml THP-1 growth media every third day. Cell lines were cultured at 37°C, 5% CO₂.

Opsonophagocytic Ab responses against HIV-1 p24 were assessed using the method of Ackerman et al. (40) with modifications to assess phagocytosis by a pDC cell line (Gen2.2). IgG was isolated from plasma using Melon gel IgG purification kits (Pierce, Rockford, IL). Plasma was diluted 1:10 with Melon gel purification buffer and subsequently mixed end over end with Melon gel purification support in a Pierce spin column for 5 min. The Pierce spin column was then centrifuged (6010 × g for 1 min), and the eluate (containing purified IgG) was collected. The concentration of IgG in the eluate was measured using a NanoDrop spectrophotometer (Thermo Scientific, Wilmington, DE) and adjusted to 1 mg/ml. Purified IgG (50 μl, 100 μg/ml) was incubated with 50 μl of a master mix containing biotinylated, recombinant (baculovirus-expressed) HIV-1 p24 (20 μg/ml; Protein Sciences, Meriden, CT) and NeutrAvidin-labeled FluoSpheres (1.8 × 10⁹ particles/ml; 0.1 μm diameter; Molecular Probes, Eugene, OR) for 2 h (37°C, 5% CO₂), in a sterile 96-well tissue culture plate. Gen2.2 cells (100 μl, 2 × 10⁵ cells/ml) were added to each well and cultured for 16 h (37°C, 5% CO₂). Gen2.2 cells were transferred to 5-ml polystyrene round-bottom tubes (BD Biosciences, San Jose, CA), washed with 1% BSA (AusgeneX, Loganholme, QLD, Australia)/PBS (Sigma-Aldrich, Castle Hill, NSW, Australia), and resuspended in the residual volume for acquisition. Data were acquired on a FACSCanto II using FACSDiva software (BD Biosciences) with an acquisition stopping gate set at 10,000 Gen2.2 cells as characterized by their forward scatter and side scatter profiles. Data files were analyzed using FlowJo software (v7.6, Tree Star, Ashland, OR).

The degree of opsonophagocytosis (phagocytic index) was calculated by multiplying the percentage of FluoSphere⁺ Gen2.2 cells with the mean fluorescence intensity of the FluoSphere⁺ Gen2.2 cells (40). Phagocytic indices were further corrected for the phagocytic index obtained when Gen2.2 cells or THP-1 cells were incubated with HIV-1 p24-conjugated beads in the absence of IgG (corrected phagocytic index). Additionally, to account for interassay variability in the results of the SCOPE cohort, each patient’s corrected phagocytic index was divided by the phagocytic index of FluoSpheres conjugated with IgG alone to produce a relative phagocytic index.

Gen2.2 and THP-1 cells (2 × 10⁵) were transferred to 5-ml polystyrene tubes, washed with 1% BSA/PBS, and stained with the following mAbs: CD64-allophycocyanin (10.1; Invitrogen, Carlsbad, CA), mouse anti-human CD32a (2C3B11B8; Sino Biological, Beijing, China), and rabbit anti-human CD32b (112; Sino Biological) for 15 min. Cells were washed with 1% BSA/PBS and the secondary polyclonal Abs, that is, goat anti-mouse IgG-Fc-PE (Ab5881; Abcam, Cambridge, MA) and goat anti-rabbit IgG-Fc-FITC (Ab98484; Abcam), were added for 30 min. Cells were washed twice with 1% BSA/PBS and resuspended in residual volume for acquisition.

Ninety-six–well microtiter plates were coated overnight at 4°C with 0.1 μg/ml HIV-1 (IIIB) p24 purified native protein (Advanced Biotechnologies, Columbia, MD) or 0.1 μg/ml HIV-1 (clade B) gp140 (Immune Technology, New York, NY). Plates were washed three times with 0.05% Tween 20/PBS and blocked with 5% BSA/PBS (1 h). Plates were washed three times and plasma serially diluted in 2% BSA/PBS was added (2 h). Plasma from an individual with high absorbance readings was used as a standard and an arbitrary unit of 1000 U/ml was assigned to the top standard. Plates were washed three times and HRP-conjugated anti-human IgG1 (Invitrogen) was diluted 1:2000 for p24 or 1:1000 for gp140 in 2% BSA/PBS and added to the plate (1 h). For both p24- and gp140-specific IgG2 ELISAs, biotinylated anti-human IgG2 (SouthernBiotech, Birmingham, AL) diluted 1:1000 in 2% BSA/PBS was added (1 h). Additionally, for IgG2, plates were further washed three times and streptavidin-conjugated HRP (BD Pharmigen, San Diego, CA) diluted 1:5000 in 2% BSA/PBS was added (1 h). Plates were washed five times and color development followed the addition of tetramethylbenzidine substrate (Sigma-Aldrich). Reactions were stopped with 1 M sulfuric acid (AnalaR grade; BDH Chemicals, Poole, U.K.) and absorbance was measured at 450 nm.

The method described by Chung et al. (41) for the depletion of IgG3 and IgG4 was modified to deplete IgG2 from purified IgG preparations. Streptavidin-conjugated M-270 Dynabeads (Invitrogen) were washed three times in PBS and incubated with biotin-conjugated mouse anti-human IgG2 (SouthernBiotech) at a mass ratio of 100:1 for 30 min (room temperature). Beads were then washed three times in 0.1% BSA/PBS and incubated with Melon gel–purified IgG samples (80:1) for 24 h (4°C). Magnetic beads were removed from the suspension using a Dynal MPC-E magnet (Dynal, Oslo, Norway) and the IgG2-depleted plasma was collected. To obtain maximal depletions, this procedure was undertaken twice for each sample. The extent of IgG1 and IgG2 depletion was assessed by ELISA using a similar protocol to that used for detection of HIV p24– and HIV gp140–specific IgG1 and IgG2 Abs, but plates were coated with 1 μg/ml recombinant protein G (Pierce) instead of Ags, and Abs were detected with either biotinylated anti-human IgG1 (Invitrogen) diluted 1:2000 or biotinylated anti-human IgG2 (SouthernBiotech) diluted 1:4000.

Statistical testing was performed using GraphPad Prism software (v5.0, GraphPad Software, San Diego, CA). Mann–Whitney t tests were performed for between-group comparisons. Wilcoxon signed rank tests were performed for comparisons between paired data sets. Correlations were assessed using a Spearman rank correlation.

To assess isotype diversification of IgG Abs to HIV-1 p24, serum levels of HIV-1 p24–specific IgG1 and IgG2 Abs in SCOPE patients were transformed using the natural log to produce a linear relationship. Robust regression was then used to test whether the relationship between log p24–specific IgG1 and log p24–specific IgG2 varied between groups (elite or viremic controllers versus noncontrollers) using an interaction between group and log p24–specific IgG2.

Initially, IgG was purified from plasma of 14 HIV controllers and 20 HIV noncontrollers (Perth cohort) and the opsonization of HIV-1 p24 conjugated to fluorescent beads and phagocytosis by the Gen2.2 or THP-1 cell lines was assessed. IgG was analyzed at four concentrations (100, 10, 1 and 0.1 μg/ml). Although opsonophagocytic Ab responses were higher with the THP-1 cell line (Fig. 1A, 1B), the dose response curves were more uniform with Gen2.2 cells (Fig. 1C, 1D). To investigate these differences between Gen2.2 and THP-1 cells, FcγR expression was assessed by flow cytometry. Whereas FcγRIIa and IIb expression was similar on the two cell lines, THP-1 cells expressed markedly higher levels of the “high-affinity” FcγRI compared with Gen2.2 cells (Supplemental Fig. 1). FcγRIIIa was not detected on Gen2.2 cells.

HIV controllers exhibited higher corrected phagocytic indices than did noncontrollers with Gen2.2 cells when IgG was incubated at a concentration of 100 (p = 0.0002) and 10 μg/ml (p = 0.0003), but this difference was lost at 1 (p = 0.07) and 0.1 μg/ml (p = 0.4) of IgG (Fig. 1E). Similarly, when the assay was performed with THP-1 cells, HIV controllers exhibited higher corrected phagocytic indices when IgG was incubated at a concentration of 100 (p = 0.0002), 10 (p = 0.0003), 1 (p = 0.007), and 0.1 μg/ml (p = 0.04) (Fig. 1F). Subsequent studies of opsonophagocytic Ab responses against HIV-1 p24 were undertaken with Gen2.2 cells alone and with IgG preparations at a concentration of 25 μg/ml.

We next attempted to replicate the findings of the prior experiment for Gen2.2 cells, but this time in a larger cohort of patients consisting of 30 elite HIV controllers, 29 viremic HIV controllers, and 30 HIV noncontrollers (SCOPE cohort). Viremic controllers exhibited higher opsonophagocytic Ab responses against HIV-1 p24 than did noncontrollers (p = 0.0001) and also elite controllers (p = 0.01). Similarly, elite controllers exhibited higher responses than did noncontrollers (p = 0.05; Fig. 2).

We next compared plasma levels of IgG1 and IgG2 Abs to HIV-1 p24 in controllers and noncontrollers from the SCOPE cohort. IgG1 Abs to HIV-1 p24 were higher in viremic controllers than in noncontrollers (p = 0.004) and elite controllers (p = 0.04), but there was no difference between elite controllers and noncontrollers (p = 0.30) (Fig. 3A). Similarly, IgG2 Abs to HIV-1 p24 were higher in viremic controllers than noncontrollers (p = 0.0004) and elite controllers (p = 0.02), but there was no difference between elite controllers and noncontrollers (p = 0.15) (Fig. 3B).

To further examine the significance of higher IgG1 and IgG2 Abs to HIV-1 p24 in viremic controllers, we also examined plasma levels of IgG1 and IgG2 Abs to HIV-1 gp140. As shown in Fig. 3C, there was no difference in plasma levels of IgG1 Abs to HIV-1 gp140 among the three groups of patients. However, plasma levels of IgG2 Abs to gp140 were higher in elite controllers compared with both viremic controllers (p = 0.003) and noncontrollers (p = 0.001; Fig. 3D).

We next sought to assess isotype diversification of HIV-1 p24–specific IgG Abs. Because a relatively higher production of downstream IgG2 Abs is indicative of greater isotype diversification of an IgG Ab response, we constructed linear regression lines by plotting plasma levels of HIV-1 p24–specific IgG1 versus those of HIV-1 p24–specific IgG2 and compared the slope of these lines between patient groups. A skewing of the slope toward IgG2 indicates a greater degree of isotype diversification. Viremic controllers exhibited greater isotype diversification of IgG Abs to HIV-1 p24 compared with noncontrollers (difference in slopes = −0.76, p = 0.009). Whereas the slope of the lines was also different, and in the same direction, for the comparison of elite controllers with noncontrollers, the difference was not statistically significant (difference in slopes = −0.51, p = 0.1). Similarly, isotype diversification was not different between viremic controllers and elite controllers (difference in slopes = −0.25, p = 0.27).

To examine functional aspects of IgG Abs to HIV-1 p24, pDC-reactive opsonophagocytic Ab responses against HIV-1 p24 were correlated with plasma levels of HIV-1 p24–specific IgG1 and IgG2 Abs and plasma HIV RNA levels in the SCOPE cohort. There were strong positive correlations of pDC-reactive opsonophagocytic Ab responses against HIV-1 p24 with both IgG1 and IgG2 Abs to HIV-1 p24 (r = 0.75, p < 0.0001 and r = 0.61, p < 0.0001, respectively; data not shown). In contrast, there was a highly significant inverse correlation of opsonophagocytic IgG Ab responses against HIV-1 p24 with plasma HIV RNA levels in all viremic patients (r = −0.51, p < 0.0001; Fig. 4A). There were also weaker inverse correlations between IgG1 and IgG2 Abs to HIV-1 p24 and plasma HIV RNA levels, moreso for IgG2 Abs (r = −0.29, p = 0.02 and r = −0.38, p = 0.0003, respectively; Fig. 4B, 4C).

To examine the possible determinants of higher opsonophagocytic Ab responses against HIV-1 p24 in HIV controllers, we correlated plasma levels of opsonophagocytic Abs and IgG1 and IgG2 Abs to HIV-1 p24 with CD4⁺ T cell counts (Table I) in the entire SCOPE cohort (n = 89). We also examined the correlation of IgG2 Abs to HIV-1 gp140 with CD4⁺ T cell counts. Opsonophagocytic Ab responses against HIV-1 p24 did not correlate with CD4⁺ T cell counts (r = −0.08, p = 0.44; data not shown). Similarly, there was no correlation between IgG1 or IgG2 anti–HIV-1 p24 and CD4⁺ T cell counts (r = −0.16, p = 0.13 and r = −0.15, p = 0.16, respectively; data not shown). However, there was a weak correlation of IgG2 anti-HIV-1 gp140 with CD4⁺ T cell counts (r = 0.23, p = 0.03; data not shown).

To determine whether phagocytosis of HIV-1 p24 opsonized by IgG Abs was mediated through FcγRIIa, Gen2.2 cells were pretreated with a mAb to FcγRIIa (AF1875, R&D Systems, Minneapolis, MN) at a concentration of 50 μg/ml or with PBS. Purified IgG (25 μg/ml) from two elite controllers, viremic controllers, and noncontrollers from the SCOPE cohort who had demonstrated the highest pDC-reactive opsonophagocytic Ab responses against HIV-1 p24 were examined. Compared to PBS pretreatment, the phagocytic indices of all samples were reduced to the level of the negative control (Gen2.2 cells incubated with HIV p24–conjugated beads in the absence of IgG) when cells were pretreated with anti-FcγRIIa (Fig. 5A).

One explanation for our observation that isotype diversification (skewing toward IgG2) of IgG Abs to HIV-1 p24 was greater in viremic controllers than elite controllers and noncontrollers is that IgG2 Abs enhance opsonization of HIV-1 p24 or phagocytosis of IgG Abs to HIV-1 p24 through FcγRIIa. We therefore depleted the Ig preparations, referred to above, of IgG2 and compared corrected phagocytic indices for those samples with the indices of samples that had not been depleted of IgG2. Following depletion of IgG2 from Ig samples by a median (range) of 86% (83–95%) (Fig. 5B) there was a small increase in phagocytosis of HIV-1 p24 opsonized by IgG Abs (p = 0.03, Fig. 5C), suggesting that IgG2 Abs did not enhance phagocytosis of opsonized HIV-1 p24 by pDCs (Gen2.2 cells).

We have demonstrated that HIV controllers, when compared with noncontrollers, exhibit an IgG Ab response against HIV-1 p24 that is characterized by higher pDC-reactive opsonophagocytic Ab activity and greater isotype diversification. Furthermore, these characteristics were higher in viremic controllers than in elite controllers. Additionally, pDC-reactive opsonophagocytic Ab responses against HIV-1 p24 inversely correlated with plasma HIV viral load in all viremic patients. We also demonstrated that IgG2 Abs to HIV-1 gp140 were higher in elite controllers than in noncontrollers and viremic controllers, supporting the findings of previous studies, which showed that slower progression of HIV disease in long-term nonprogressors was associated with IgG2 Abs against HIV Env Ags (33, 34).

Numerous studies have demonstrated that higher serum/plasma levels or avidity of IgG Abs against HIV-1 Gag proteins are associated with slower progression of HIV disease (reviewed in Ref. 11). However, to our knowledge, this is the first study to show an association between the functional activity of an IgG Ab response against HIV Gag proteins and natural control of HIV infection. It has been proposed that IgG Abs against HIV-1 p24 are markers of CD4⁺ T cell numbers and/or responses against Gag proteins (42). However, we did not demonstrate a correlation between CD4⁺ T cell counts and either opsonophagocytic Ab responses against HIV-1 p24 or the plasma level of IgG1 and IgG2 Abs to HIV-1 p24. Furthermore, our observation that pDC-reactive opsonophagocytic Ab responses against HIV-1 p24 were negatively correlated with plasma HIV RNA levels in all viremic patients provides evidence that these Abs might contribute to control of HIV replication.

A notable finding of our study was that IgG Ab responses against HIV-1 p24 were higher in viremic controllers than in elite controllers as well as noncontrollers. This finding was demonstrated by independent assays (ELISAs and pDC-reactive opsonophagocytic Abs) and corroborated by the finding that isotype diversification of IgG Abs against HIV-1 p24 was higher in viremic controllers than in elite controllers, who had a greater diversification than did noncontrollers (Fig. 3E). Possible explanations for these findings are, on the one hand, that higher Ab responses are a consequence of greater HIV replication in viremic controllers compared with elite controllers, and, on the other hand, that IgG Abs to HIV-1 p24 contribute to control of HIV replication in patients who cannot exert elite control of the infection. We posit two arguments against the former explanation. First, patients with the highest amount of HIV replication (noncontrollers) had the lowest magnitude and isotype diversification of IgG Ab responses against HIV-1 p24, which is unlikely to simply reflect more severe immunodeficiency in this group because there was no relationship between IgG Ab responses against HIV-1 p24 and CD4⁺ T cell counts in the whole SCOPE cohort, which was selected for patients with CD4⁺ T cell counts of >350/μl. Second, IgG2 Abs to HIV gp140 were highest in elite controllers who, by definition, had the lowest amount of HIV replication.

Our findings for opsonophagocytic Ab responses and plasma levels of IgG1 and IgG2 Abs against HIV-1 p24 are similar to those for CD4⁺ T cell responses against HIV-1 Gag proteins in HIV patients with persistent low-level HIV replication on ART, who exhibited higher CD4⁺ T cell responses against HIV-1 Gag proteins than did ART-treated patients with optimally suppressed HIV replication or ART-naive patients (43, 44). In those studies, the highest responses were also demonstrated in individuals with low but detectable HIV viremia. Theoretically, in states in which the immune system is functional and contributing to the control of a pathogen (as in HIV controllers), a higher steady-state level of that pathogen (HIV RNA in this case) will be associated with higher host responses aimed at controlling that pathogen (IgG Ab responses) (43, 44). This proposal would provide an explanation for why other investigators have not shown an association between IgG1 or IgG2 Abs to HIV-1 Gag proteins and control of HIV infection in unselected HIV controllers (45) or HIV controllers who did not carry protective HLA-B alleles (46), because those studies examined predominantly, or only, elite HIV controllers.

Our hypothesis that pDC-reactive opsonophagocytic Ab responses against HIV Gag proteins contribute to control of HIV infection is supported by the findings of studies on influenza virus infection in mice, which demonstrated that nucleoprotein-specific IgG Abs mediated clearance of, and heterotypic immunity against, influenza viruses (47). Clearance of influenza virus infection was dependent on class-switched IgG Abs against influenza virus nucleoproteins that mediated downstream IFN-α/β production by a mechanism involving the cooperation of FcγR and TLR7 (48). We propose that a similar control mechanism, involving an opsonophagocytic Ab response against complexes of HIV-1 Gag proteins and HIV RNA, their uptake by pDCs, and subsequent downstream activation of TLR7 and IFN-α–mediated restriction of viral replication, is a potential control mechanism for HIV-1 infection. Although it is possible that production of IgG Abs to HIV-1 Gag proteins is a consequence of CD8⁺ T cell– or NK cell–mediated killing of HIV-1–infected cells and subsequent release of HIV Gag/RNA complexes contained in their cytosol (49, 50), a pDC-reactive opsonophagocytic Ab response against HIV-1 Gag/RNA complexes may be a complementary immune control mechanism to CD8⁺ T cell responses, as suggested by studies of influenza virus infection in mice (47).

There is a growing awareness that the function of IgG Abs determined by the hinge and Fc regions of IgG molecules is crucial to the activity of some IgG Ab responses against HIV-1 Ags, and that this reflects the IgG subclass composition of Abs and affects Ab-mediated activation of effector cell responses, such as those mediated by NK cells via FcγRIIIa and/or FcγRIIc (51) and phagocytic cells via FcγRIIa (52). Ab-dependent NK cell responses against HIV Env are higher in elite HIV controllers than noncontrollers (53), and prevention of HIV-1 infection in subjects enrolled into the RV144 HIV vaccine study has been associated with both Ab-dependent NK cell responses and production of IgG3 Abs against Env (41). Our study, which addressed the role of Abs in the control rather than prevention of HIV infection, demonstrated that opsonophagocytic Ab responses against HIV-1 p24 mediated by pDCs (Gen2.2 cells) occurred through FcγRIIa (Fig. 5A), which is the major activating FcγR on human pDCs (31). Possession of the FcγRIIa^H-131 genotype, which encodes for an FcγRIIa variant with high-affinity binding to IgG Abs (28), is associated with a slower rate of CD4⁺ T cell decline in HIV-1–infected men (54). Additionally, we have shown that HIV-infected patients who possessed both the FcγRIIa^H-131 genotype and produced IgG2 Abs against HIV-1 p24 after vaccination with a DNA vaccine encoding HIV Gag-Pol and IFN-γ maintained lower viral loads than did those patients who had only one of these characteristics (55).

We also demonstrated that isotype diversification (skewing toward IgG2) of IgG Abs to HIV-1 p24 was greater in viremic HIV controllers than in elite controllers and noncontrollers. Furthermore, although plasma levels of IgG1 or IgG2 Abs to HIV-1 p24 exhibited weaker negative correlations with plasma HIV RNA than did pDC-reactive opsonophagocytic Ab responses against HIV-1 p24 in all viremic HIV patients, the correlation for IgG2 anti–HIV-1 p24 was stronger than for IgG1 anti–HIV-1 p24 (Fig. 4). These findings suggest that enrichment of the IgG Ab response against HIV-1 p24 for IgG2 Abs is a component of the IgG Ab response against HIV-1 Gag that is associated with control of HIV infection. We therefore examined the effect of IgG2 Abs in the assay of pDC-reactive opsonophagocytic Ab responses against HIV-1 p24. Depletion of IgG2 from Ig preparations slightly increased phagocytosis of HIV-1 p24–coated beads complexed with Abs. These findings support those of Forthal et al. (56), who demonstrated that IgG2 Abs inhibited phagocytosis of Ab-opsonized HIV-1 virus-like particles by monocytes, and are in accord with observations that IgG2 has a lower affinity for all Fcγ receptors than do other IgG subclasses (28). Therefore, whereas the function of the Fc region of IgG2 Abs is more restricted to phagocytosis than other IgG subclasses, by virtue of dominant binding to FcγRIIa^H-131 (28), our findings suggest that IgG2 Abs are likely to exert their effect on opsonophagocytic Ab responses against Ags of the HIV core by affecting opsonization and/or immune complex formation more than Fc receptor binding. This may reflect the characteristics of IgG2 molecules that facilitate the binding of the Fab region to multivalent Ags (23–27). Furthermore, plasma IgG/IgM complexes containing IgG2 predominate in healthy individuals (57) and we have shown that FcγRIIa-binding immune complexes in the plasma of HIV controllers are enriched for IgG2 when compared with noncontrollers (35).

As with all studies of immune correlates of HIV-1 control, we acknowledge that our study is unable to discern cause and effect. However, our data present evidence that IgG Ab responses against HIV-1 p24 may contribute to control of HIV replication and of a mechanism by which this may occur. We also acknowledge that by using the Gen2.2 cell line to assay opsonophagocytic Ab responses we have not addressed the possible contribution of the patient’s pDCs. However, in this study, we sought to analyze the opsonophagocytic function of IgG Abs against HIV-1 p24 independent of cellular factors.

In summary, we provide novel data indicating that natural control of HIV-1 infection is associated with higher pDC-reactive opsonophagocytic Ab responses and greater isotype diversification of IgG Abs against HIV-1 p24 compared with noncontrollers. Furthermore, this effect was most notable in viremic controllers and was independent of CD4⁺ T cell counts. Further studies are required to elucidate the properties of IgG2 Abs against HIV proteins that are associated with natural control of HIV-1 infection.

We thank Drs J. Plumas and L. Chaperot for providing the Gen2.2 cell line.

The authors have no financial conflicts of interest.